On the Analysis of Training Data for WaveNet-Based Speech Synthesis

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F49777513%3A23520%2F18%3A43952771" target="_blank" >RIV/49777513:23520/18:43952771 - isvavai.cz</a>

Výsledek na webu

<a href="https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8461960" target="_blank" >https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8461960</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/ICASSP.2018.8461960" target="_blank" >10.1109/ICASSP.2018.8461960</a>

Jazyk výsledku

angličtina

Název v původním jazyce

On the Analysis of Training Data for WaveNet-Based Speech Synthesis

Popis výsledku v původním jazyce

In this paper, we analyze how much, how consistent and how accurate data WaveNet-based speech synthesis method needs to be able to generate speech of good quality. We do this by adding artificial noise to the description of our training data and observing how well WaveNet trains and produces speech. More specifically, we add noise to both phonetic segmentation and annotation accuracy, and we also reduce the size of training data by using a fewer number of sentences during training of a WaveNet model. We conducted MUSHRA listening tests and used objective measures to track speech quality within the conducted experiments. We show that WaveNet retains high quality even after adding a small amount of noise (up to 10%) to phonetic segmentation and annotation. A small degradation of speech quality was observed for our WaveNet configuration when only 3 hours of training data were used.

Název v anglickém jazyce

On the Analysis of Training Data for WaveNet-Based Speech Synthesis

Popis výsledku anglicky

In this paper, we analyze how much, how consistent and how accurate data WaveNet-based speech synthesis method needs to be able to generate speech of good quality. We do this by adding artificial noise to the description of our training data and observing how well WaveNet trains and produces speech. More specifically, we add noise to both phonetic segmentation and annotation accuracy, and we also reduce the size of training data by using a fewer number of sentences during training of a WaveNet model. We conducted MUSHRA listening tests and used objective measures to track speech quality within the conducted experiments. We show that WaveNet retains high quality even after adding a small amount of noise (up to 10%) to phonetic segmentation and annotation. A small degradation of speech quality was observed for our WaveNet configuration when only 3 hours of training data were used.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

20205 - Automation and control systems

Projekt

Výsledek vznikl pri realizaci vícero projektů. Více informací v záložce Projekty.

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2018

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

ISBN

978-1-5386-4658-8

ISSN

—

e-ISSN

2379-190X

Počet stran výsledku

5

Strana od-do

5684-5688

Název nakladatele

IEEE

Místo vydání

New York

Místo konání akce

Calgary, AB, Canada

Datum konání akce

15. 5. 2018

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

000446384605169